smile.classification

Class NeuralNetwork.Trainer

java.lang.Object

smile.classification.ClassifierTrainer<double[]>

smile.classification.NeuralNetwork.Trainer

Enclosing class:

NeuralNetwork

public static class NeuralNetwork.Trainer
extends ClassifierTrainer<double[]>

Trainer for neural networks.

Constructor Summary

Constructors

Constructor and Description
Trainer(NeuralNetwork.ErrorFunction error, int... numUnits) Constructor.
Trainer(NeuralNetwork.ErrorFunction error, NeuralNetwork.ActivationFunction activation, int... numUnits) Constructor.

Method Summary

Modifier and Type	Method and Description
NeuralNetwork.Trainer	setLearningRate(double eta) Sets the learning rate.
NeuralNetwork.Trainer	setMomentum(double alpha) Sets the momentum factor.
NeuralNetwork.Trainer	setNumEpochs(int epochs) Sets the number of epochs of stochastic learning.
NeuralNetwork.Trainer	setWeightDecay(double lambda) Sets the weight decay factor.
NeuralNetwork	train(double[][] x, int[] y) Learns a classifier with given training data.

Methods inherited from class smile.classification.ClassifierTrainer

setAttributes

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Trainer

public Trainer(NeuralNetwork.ErrorFunction error,
               int... numUnits)

Constructor. The activation function of output layer will be chosen by natural pairing based on the error function and the number of classes.

Parameters:

error - the error function.

numUnits - the number of units in each layer.

Trainer

public Trainer(NeuralNetwork.ErrorFunction error,
               NeuralNetwork.ActivationFunction activation,
               int... numUnits)

Constructor.

Parameters:

error - the error function.

activation - the activation function of output layer.

numUnits - the number of units in each layer.

Method Detail

setLearningRate

public NeuralNetwork.Trainer setLearningRate(double eta)

Sets the learning rate.

Parameters:

eta - the learning rate.

setMomentum

public NeuralNetwork.Trainer setMomentum(double alpha)

Sets the momentum factor.

Parameters:

alpha - the momentum factor.

setWeightDecay

public NeuralNetwork.Trainer setWeightDecay(double lambda)

Sets the weight decay factor. After each weight update, every weight is simply ''decayed'' or shrunk according w = w * (1 - eta * lambda).

Parameters:

lambda - the weight decay for regularization.

setNumEpochs

public NeuralNetwork.Trainer setNumEpochs(int epochs)

Sets the number of epochs of stochastic learning.

Parameters:

epochs - the number of epochs of stochastic learning.

train

public NeuralNetwork train(double[][] x,
                           int[] y)

Description copied from class: ClassifierTrainer

Learns a classifier with given training data.

Specified by:

train in class ClassifierTrainer<double[]>

Parameters:

x - the training instances.

y - the training labels.

Returns:

a trained classifier.